The commercial conditions for getting a set of records into an aggregated search environment seem to often be more onerous than allowing a broadcast search of the native source. Part of the problem lies in perception of the aggregated environment as being a competitor to the source of the records. This may be a true perception, a false one, or a wishful one.

I think the answer is a hybrid service. Index what you can (discovery service), federate what you can, harvest what you can, etc. A good federated search framework can integrate access to content that has been obtained by a number of different technologies.

In other words, lead with the question of what sources are important not with the question of what technology is the most convenient. Then find ways to get access to each of them. I’ve never said that federated search makes discovery services or crawling or harvesting obsolete.

With either broadcast search or an aggregated index, you only have access to what you have access to. And in either case, providing access to any arbitrary resource may or may not be possible, and will take a non-zero amount of ‘development’ time to add. With either case, the user using the tool only has access to searching content included in the tool, unless they leave the tool, which they can do in either case.

It seems to me you are assuming that broadcast search will have a lower barrier to including additional resources, and end up having more coverage? This to me is something that needs to be seen in practice, not assumed theoretically. The key thing will be comparing two actual tools, and seeing which tool includes more resources, and more relevant and quality resources, for some particular context (audience and use cases).

If SerSol’s aggregated index comes up short, then that will be to it’s detriment. But I don’t see how that can be assumed without an actual evaluation of two actual deployed tools side by side.